اختلالات اضطراب سلامت: تفسیر شناختی

The features of severe health anxiety, intense and persistent anxiety about one's present and future health, are described. In common with other anxiety disorders such as GAD, PTSD and OCD, the core of HAD is distressing, uncontrollable anxiety, and is classifiable as an Anxiety Disorder, Health Anxiety Disorder (HAD). The cognitive construal of HAD proposes that health anxiety is caused by catastrophic misinterpretations of the significance of sensations and/or changes in bodily functions and appearance (such as swellings, pain, loss of energy, dizzy spells). The nature, causes, triggers, persistence, assessment and treatment of HAD are reviewed, and the present status of the cognitive model is appraised. Suggestions are made for future research and clinical applications, and the need for incisive evaluations of the main premises of the model is emphasized.

The phenomenon of health anxiety described by Salkovskis and Warwick in 1986 provided a foundation for the concept of Health Anxiety Disorders, which is now classifiable as an anxiety disorder in association with PTSD, OCD, SAD, PD and GAD ( Noyes, 1999; Salkovskis & Warwick, 1986). The common element that links the various anxiety disorders is the presence of excessive anxiety. In health anxiety the excessive anxiety is focused on one's present and future health. Fig. 1 Full-size image (27 K) Fig. 1. The relation of health anxiety disorders to other types of anxiety disorder. Figure options Salkovskis (1989; Salkovskis & Warwick, 2001) proposed that health anxiety is caused by catastrophic misinterpretations of sensations and symptoms. In common with all of the anxiety disorders, health anxiety is construed as occurring along a continuum from minimal to excessive anxiety (and clinically significant). The aim of the present construal is to collect and arrange disparate information and ideas pertaining to health anxiety, including the nature, causes, persistence, safety behaviour, assessment and treatment of HAD. The application of a cognitive approach to construing health anxiety was a crucial step. “The cognitive theory states that in severe health anxiety…bodily signs, symptoms, variations, and medical information tend to be perceived as more dangerous than they really are, and that a particular illness is believed to be more probable than it really is” (Salkovskis, 1996, p. 65). The tendency grossly to overestimate the probability of becoming ill and the seriousness of the dreaded illness, is persistent. “Anxiety focused upon health is an almost universal phenomenon…and persistent anxiety about health is common both in the community and in the clinic” (Salkovskis & Warwick, 2001, p. 46). The anxiety is provoked by perceived changes in bodily appearance or functioning, exposure to negative information about health that is personally significant and/or by experiences that are catastrophically misinterpreted (e.g., disturbing intrusive images). The sensitivity to changes in bodily functions and appearance tends to be heightened during an illness. Notably, health anxiety is not suspended when a person is actually ill, and the overestimations of the probability and seriousness of threats to one's health continue to operate (Salkovskis & Warwick, 2001). The present construal arose out of the infusion of cognitive concepts into clinical psychology, and emerged from the success of the cognitive model of panic formulated by Clark (1986), and Salkovskis's (1985) cognitive model of obsessive compulsive disorders (OCD). The essence of Clark's model is that episodes of panic are caused by a catastrophic misinterpretation of certain bodily sensations, such as a pounding heart. ‘My pounding heart means that I am about to have a heart attack’. The danger is imminent. The panic model did not address the catastrophic misinterpretations of sensations or signs that signal a future danger to one's health, such as developing cancer. This important difference was a spur to the formulation of the health anxiety concept. Salkovskis's (1985) cognitive analysis of OCD led him to connect important features of that anxiety disorder with serious fears of dangers to one's health. The cognitive model of health anxiety absorbed some aspects of hypochondriasis, and then expanded the definition, scope, classification, nature and treatment of severe health anxiety. Health anxiety disorder is a positive ‘diagnosis’ or rather, a positive classification of a psychological problem within the existing group of anxiety disorders. Hypochondriasis Severe health anxiety, the extreme end of a continuum of health anxiety, is often termed hypochondriasis. At times the two terms, hypochondriasis and severe health anxiety, are used interchangeably, but a clear distinction would be helpful. ‘Hypochondria’ is an anatomical term. It describes “those parts of the human abdomen…which lie immediately under the ribs… [where] the viscera [are] situated… the liver, gall-bladder, spleen etc.” (Compact Oxford English Dictionary, 1971, p. 507). They were believed to be the source of melancholy fumes “low spirits for which there is no real cause” (p. 507). Over time the meaning of hypochondria expanded into “a disorder of the nervous system…chiefly characterized by the patient's unfounded belief that he is suffering from some serious bodily disease” (p. 597). Hypochondria is currently classified as a mental disorder (DSM-IV, 2005; see Asmundson, Taylor, and Cox (2001) for an account of the various iterations of ‘hypochondria’ in the DSM). The diagnosis of hypochondriasis is categorical: a distorted belief that one is suffering from a serious disease despite all the medical and other evidence against the belief; a disease conviction. Hypochondriacal beliefs are resistant to disconfirmation. Unlike health anxiety, in which future dangers are anticipated, in hypochondriasis the danger is present and active, and the belief is fixed. The diagnosis of hypochondriasis has been criticised (e.g., Abramowitz & Moore, 2007; Asmundson et al., 2001; Creed & Barsky, 2004; Deacon & Abramowitz, 2008; Olatunji, Deacon, & Abramowitz, 2009; Schmidt, 2004; Warwick & Salkovskis, 1990; Wells & Hackmann, 1993; Wise & Birket-Smith, 2002). The cognitive model The cognitive construal of health anxiety has promoted a substantial increase in knowledge about fears pertaining to one's current state of health and future health (Asmundson et al., 2001; Marcus, Gurley, Marchi, & Bauer, 2007; Salkovskis & Warwick, 2001; Warwick & Salkovskis, 1990). Attention has been paid to the cognitions involved in anxiety about one's health. The inflation of vigilance and the triggers for heightened anxiety have been incorporated in the model. The concept of health anxiety now leads to the inclusion of fears of mental illness as well as physical illness. A good deal of attention has been paid to the operation of cognitive biases (such as ex consequentia reasoning; Arntz, Rauner, & van den Hout, 1995; thought–action–fusion; Shafran & Rachman, 2004; confirmatory biases; Salkovskis & Warwick, 2001; the overprediction of fear; Rachman & Bichard, 1988; overprediction of pain; Rachman & Arntz, 1991). Given the recent reconsideration of the nature of safety behaviour ( Rachman, Radomsky, & Shafran, 2008), the role of the compelling search for reassurance in patients with health anxiety ( Salkovskis & Warwick, 2001) has been expanded to include positive safety behaviour. Arising out of the cognitive model of OCD the powerful effects of intrusive images in health anxiety are now under investigation (see below). Specific forms of cognitive behaviour therapy ( Clark et al., 1998; Sorensen, Birket-Smith, Wattar, Buemann, & Salkovskis, 2011; Warwick, Clark, Cobb, & Salkovskis, 1996) have been deduced from the model and are being subjected to controlled evaluations. Evaluating the effects of therapy can be confusing when the terms hypochondriasis and health anxiety are used interchangeably (e.g., Clark et al., 1998; Nakao, Shinozaki, Ahern, & Barsky, 2011; Sorensen et al., 2011; Visser & Bouman, 2001; Warwick et al., 1996). The criteria for selecting participants vary from trial to trial (hypochondriasis or severe health anxiety), as do the methods of assessment and the treatment protocols. However, it should be possible to avoid some of the obstacles that have impeded collection of data about prevalence. HAD is dimensional and a range of promising psychometric and other methods of assessment are available or in progress of development (see below). HAD cognitions The affected people are morbidly preoccupied with their health and greatly overestimate the probability that they have a serious illness and/or that they are at risk of developing a serious illness. They also overestimate the seriousness of the dreaded illness; they fear that it will be extremely disabling and not infrequently people fear that it will be fatal. Changes in bodily sensations, functions, appearance, and unwanted disturbing intrusive thoughts/images are misinterpreted as indications of an actual or potentially serious illness, possibly fatal. They believe that they are exceptionally vulnerable to illnesses in general and/or exceptionally vulnerable to a particular illness (e.g., cancer, but are relatively unconcerned about cardiac illnesses). The particular, circumscribed fears are idiosyncratic and originate from the experience of a distressing illness or treatment, or a distressing illness/treatment of a relative or friend. The common fears are a dread of cancer, heart attacks, AIDS, strokes, mental illness. Recognition that anxiety about developing a mental illness is a form of health anxiety will ultimately provide help for sufferers, who are likely to benefit from improved diagnoses and from CBT that is used to treat health anxiety. The recent development of scales to assess cognitions about one's health ( Hadjistavropolous et al., in press) will enhance our knowledge about the content and significance of the cognitions, and most importantly, facilitate experimental evaluations of the cognitive model. In addition to these personally-relevant health cognitions, there are indications that people with HAD also have anxiety-evoking general beliefs about health and illness, such as there is far more illness in the world than people realise, there must be an explanation for all aches, pains, unusual sensations, and so forth ( Fulton, Marcus, & Merkey, 2011). In the cognitive construal it is the person's (health) cognitions that are critical; that is, the interpretation which the person places on experiences, events, information. These interpretations are personally significant. Catastrophic misinterpretations are particularly damaging. General beliefs, held by many other people, are not personally significant, and may play little or no part in HAD. What causes health anxiety? There are three pathways to the acquisition of fears. They are acquired by experiencing disturbing or damaging events, by vicarious acquisition (such as a serious illness experienced by a relative or close friend), or by absorbing threatening information (Rachman, 1990). An example of the vicarious acquisition of an intense, specific fear was seen in the case of a middle-aged woman with an unremarkable medical history who was tormented through most of her adult life by a fear that she was going to suffer a catastrophic stroke that would leave her severely incapacitated. At the age of 10 her mother had a sudden stroke that left her with a mild speech impediment and a disfiguring weakness on the left side of her face. The patient was disturbed by recurrent images of herself with a similar disfigurement. In order to protect herself she tried to avoid any emotion-provoking situations, restricted her diet by excluding salty or spicy foods, or alcohol, measured her blood pressure several times a day, and compulsively checked the appearance of her face. She was little concerned about other threats to her health. Another patient was extraordinarily preoccupied with potential threats to her health and spent several hours every day reading and re-reading the Merck Medical Manual and scanning the internet for information about health risks. She was so anxious about her health that she maintained a restricted, hypervigilant lifestyle, feeling that at any time she could become seriously ill. She avoided travel, kept to a strict narrow diet, and avoided going anywhere near sick people or hospitals. Following a bout of ill-health and a lengthy period of asthma during childhood she regarded herself as a weak and vulnerable person, and indeed for approximately three years had been advised to refrain from taking part in sports or other strenuous activities at or after school. At the time of her psychological assessment in adulthood it was affirmed that her health was not compromised and numerous medical tests and consultations were all negative. The fears were acquired by a lengthy period of adverse health and her feelings of vulnerability were reinforced by her extremely overprotective parents. Fears of becoming mentally ill can be acquired vicariously by proximity to people who appear to be mentally disturbed. They are avoided in public places, and psychiatric hospitals and wards are strictly off-limits. In one case the affected man's fear of becoming mentally ill was so intense that each working day he took a 4 mile roundabout route to work in order to avoid seeing the local psychiatric hospital. The fear had been generated when he was boy, and his father insisted on taking him on monthly visits to his severely ill uncle who was a longtime resident in a psychiatric care-home, and rarely spoke or responded to his visitors. Some patients with health anxiety resort to explaining their fear of proximity to disturbed people in terms of contracting ‘mind germs’ which can affect them. Incidentally, notions of this kind are expressed by people who are not ignorant or deluded. The idea is that mind germs share some properties of disease germs, including contagiousness. Fears of losing control of one's mind, of carrying out violent actions, ending up in a back ward of a long stay psychiatric hospital, are examples of frightening threats to a person's mental health. Perceived threats to one's sanity are a manifestation of health anxiety and can be as distressing and disabling as perceived threats to one's physical health. There are no precise data on the matter at present but a minority of patients with health anxiety, and a minority of patients with OCD, fear that they are vulnerable to mental illness, and both groups avoid proximity to people who behave in a bizarre manner in public or are known to be suffering from a mental illness. The underlying fear is that mental illness might be contagious. Intense fears of a threat to one's health can be produced by the absorption of disturbing information about epidemics (e.g., SARS, sudden respiratory illness) or by illnesses that can be contracted by contact with sufferers (e.g., AIDS), or by information about potentially harmful foods (e.g., carcinogens). There are several sources of potentially frightening information; friends or relations, the media, internet surfing, advertising. People who are particularly sensitive to threatening information are vulnerable to generalized fears of illness. From a treatment perspective the specific, idiosyncratic cases of health anxiety are probably more manageable than the generalized forms. The predisposing factors that underlie generalized health anxiety are not clearly delineated but high levels of anxiety sensitivity, neuroticism, depression, and a preoccupation with health and well-being are probably involved (McLure & Lilienfeld, 2001; Williams, 2004). There is evidence of elevated scores on scales of neuroticism in people with health anxiety (Williams, 2004), but little to suggest a unique or even specific contribution of neuroticism. In research on vulnerability the dependence on information collected retrospectively is a limitation. The feelings of fragility, emotional and/or physical, described by patients with HAD certainly require investigation and longitudinal studies are needed. Triggers of health anxiety Bodily sensations, changes in appearance (swellings, blemishes etc.), injuries, changes in bodily functioning (such as loss of energy, insomnia, loss of memory, fatigue), alarming information about health risks, intrusive images/thoughts, proximity of perceived environmental threats (e.g., blood, putrefying foods), proximity to evidently sick people, thoughts of losing control and behaving in a bizarre manner, episodes of depression, proximity to mentally disturbed people, news of distressing illnesses of relatives/friends, episodic and/or chronic pain, recurrent intrusive images about pain and its consequences. Pain is intrusive, frequently distressing, dominating, and often is an alarming symptom. It is a very common reason for seeking medical assistance. Mantyselka et al. (2001) found that 40% of 5646 patients attending a primary care facility had pain complaints, and in a Swedish study the figure was 28% of 6890 patients (Hasselstrom, Liu-Palmgren, & Rasjo-Wraak, 2002). High levels of health anxiety are common in patients with chronic pain; 59% of 161 patients (Rode, Salkovskis, Dowd, & Hanna, 2006). Similarly, Hadjistavropoulos, Owens, Hadjistavropolous, and Asmundson (2001) found that patients attending a clinic for pain problems had significantly elevated levels of health anxiety. Patients suffering from chronic pain and high health anxiety engage in more safety behaviours than do patients with comparable pain but low health anxiety and the resort to safety behaviours was correlated with catastrophic thoughts about the pain (Tang et al., 2007). Many sufferers from chronic pain report frequent and distressing intrusive images that are comparable to those described by patients with health anxiety. In a study carried out on 59 attendees at a Rehabilitation Clinic, Philips (2011) found that pain-related images evoked strong emotional reactions which included elevated anxiety. The images were assessed by structured interviews and experimental probes, and 78% of the participants reported one or more recurrent images when in pain. When asked to form their most distressing (index) image, their pain levels rose and negative emotions increased. In addition, their negative cognitive appraisals worsened during and after the image formation. Most of the images were ‘forward’ and involved dreaded long term consequences of the chronic pains (e.g., incapacitation, severe disablement). Before taking part in the study most of the participants were unaware of the recurrency of their pain-related images. When they described the images many had a strong emotional reaction within minutes or even seconds. In a neat experimental investigation Abramowitz and Moore (2007) demonstrated that when 27 patients diagnosed with hypochondriasis were exposed to personally significant health-anxiety triggers, they promptly experienced large increases in anxiety. The use of personally significant triggers, and personally significant safety behaviour is a strength of this incisive experiment. The “prevalence of physical symptoms within the American population at any given time is remarkably high” according to Pennebaker (1982, p. 5), and he was able to compile a list of 54 common symptoms, including headaches, coughing, nasal congestion, constipation, toothache, dizziness, back pain, muscle soreness. The formidable list of potential triggers of health anxiety is consistent with this high base rate of symptoms and signs. Notwithstanding the remarkably high prevalence of symptoms and potential triggers of alarm, we seem to manage tolerably well. As Shelley Taylor and Brown (1988) explained, most people are unrealistically optimistic about their health. This raises an intriguing question about the continuum of health anxiety. If physical, and mental, symptoms are so common, and exposures to worrying external and internal information and signs are so ubiquitous, we need to understand how people find the time to cope with potential health fears. Taylor and Brown (1988) suggest that the unrealistic optimism about one's health is an illusion. If so, it is one of the effective illusions. Maintaining factors Why does unadaptive behaviour persist? This question arises in all analyses of anxiety disorders and was termed the “neurotic paradox” by Mowrer (1948). Anxious and distressing images and thoughts about one's health, and associated behaviour, are unadaptive but they persist. The thoughts and fears persist despite the negative results of medical tests, and the provision of repeated reassurances. Clark's (1999) account of the factors that maintain anxiety disorders provides a framework for the present analysis. The belief that one's health is under serious threat produces hypervigilance; attention is attuned to signs of threat (Eysenck, 1992): to triggers. The active search for signs of threat ‘succeeds’ in detecting dangers, and the successes inflate the person's anxiety. This in turn reinforces the need for vigilance and continuous searching. The content of the fear directs the person's search for safety. Affected people are attuned to the triggers that pertain to their particular fear and scan their social, physical and internal environments for signs of danger; they are seldom ‘off-duty’. Health anxiety is persistent and pervasive. Memories of threatening information are available, emerge spontaneously, are recalled deliberately, and even ‘rehearsed’ as in post-event processing of upsetting social events (Clark & Wells, 1995; Rachman, Gruter-Andrew, & Shafran, 2000). The deliberate recall and replaying of events or information that appear to threaten one's health (post-event processing) can contribute to the maintenance of health anxiety. Memories of these threats can also recur in the form of disturbing unwanted intrusive images that evoke considerable fear. Other types of intrusive images, which involve future dreads rather than recollections, are common in cases of health anxiety (Muse, McManus, Hackmann, Williams, & Williams, 2010). In their interview study of 55 patients diagnosed with hypochondriasis, the large majority of participants reported experiencing recurrent intrusive and distressing images… “the majority of the patients (86.05 %) classified their image as relating to the future. Imagining a future event has been shown to increase an individual's perception that that the imagined event will occur” (Muse et al., 2010, p. 7). For example, ‘I have a recurrent image of myself severely disabled by the illness and unable to feed, clean or care for myself’. The characteristic response to these disturbing images is to suppress them, but unfortunately these attempts often go wrong and increase rather than decrease the frequency of the images (Rassin, 2005), thereby adding to the anxiety. In sum, several factors contribute to the maintenance of health anxiety: hypervigilance, enhanced attention to threats, especially the selective attention to threatening information about health risks, an accumulation of threatening memories, disturbing intrusive images, post-event processing, cognitive biases such as ex consequentia reasoning, and the resort to safety behaviour which reinforces the maladaptive interpretations of certain bodily sensations, changes in bodily functioning, pains, changes in appearance, intrusive thoughts and images. Intrusive images The clinical significance of intrusive images was recognised by Beck (1976) and by Wolpe (1958) but they used images in different ways. “In employing systematic desensitization, for instance, I customarily ask for a detailed description of each image. The patient's report is often very informative and, on many occasions, reveals new problems that had not been previously identified” (Beck, 1976, p. 221). “Many times, maladaptive ideation occurs in a pictorial form instead of, or in addition to, the verbal form” (Beck, 1976, p. 242). He then went on to help the patient modify the images by ‘decentering’ or ‘distancing’ them; an early, prescient example of rescripting. Wolpe (1958) on the other hand used specially prepared lists of anxiety-provoking images as an effective and economical method for exposing his patients to a full range of manipulable stimuli. The purpose of the exposures was to reduce the patient's reactions to the images, not to change them: to desensitize them. Until recently research on OCD was confined to recurrent, disturbing thoughts and impulses but the third form of recurrency, images, was neglected ( Rachman & Hodgson, 1980). However, in 1999 de Silva and Marks provided an account of intrusive images in this disorder, and Speckens, Hackmann, Ehlers, and Cuthbert (2007) reported that 81% of 37 patients with OCD experienced intrusive images. Given some of the similarities between OCD and health anxiety, it is to be expected that significant intrusive images are experienced by sufferers from health anxiety. Intrusive images in OCD can be powerful and provoke intense emotional reactions. They are primarily visual, usually vivid, effortless, fully formed, of short duration (because the person ‘stops’ the image), and the content is remarkably stable (Rachman, 2007). Some of them remain unchanged for a lifetime. Wells and Hackmann (1993) described a pilot study of 10 patients with health anxiety who were selected because they had intrusive images, and observed that the main themes were illness and death, and many were forward rather than past images. As mentioned, Muse et al. (2010) found that a large majority of hypochondriacal patients experienced “recurrent, distressing intrusive images” (p. 1). The most significant (index) image was experienced 3.77 times per week. The majority of the images were either a memory of an event or associated with a memory, but were future-oriented, and almost all were on the theme of serious illness/death. Some of their findings resonate with prominent aspects of obsessive compulsive disorders “participants reported responding to experiencing intrusive images by engaging in avoidance, checking, reassurance-seeking, distraction and rumination” (Muse et al., 2010, p. 1). As mentioned above, intrusive pain-related images are common and usually evoke negative reactions that include anxiety (Philips, 2011). Unwanted intrusive images and thoughts can contribute to health anxiety especially for those patients whose preoccupying fear is that they are developing, or might develop, a serious mental illness (Whittal, Woody, McLean, Rachman, & Robichaud, 2010). Most of the current theorising about and research into health anxiety is confined to physical illnesses, notably cancer and cardiac illness, to the neglect of threatening mental illnesses. The recurrency of the images and thoughts stirs up anxiety, and leaves the person vulnerable to persisting anxiety. Muse et al. (2010) raise the plausible possibility that these recurrent future-oriented images of illness and death “serve to maintain anxiety about health by increasing the participants' estimation of the likelihood of these events occurring” (p. 7). This resembles the effects of the cognitive bias thought–action–fusion encountered in cases of OCD (Shafran & Rachman, 2004). Recurrent intrusive thoughts of feared misfortunes (e.g., losses, accidents or illnesses) are believed by the affected person to increase the probability of the misfortune actually occurring. Muse et al. recommend research into the potential value of using rescripting (see Hackmann, Bennet-Levy, & Holmes, 2011) as part of the treatment of health anxiety, and to this may be added the modification of any thought–action–fusion cognitive biases that are encountered. The modification of thought–action–fusions can be relatively straightforward with non-clinical participants (Marino-Carper, Negy, Burns, & Lunt, 2010; Zucker, Craske, Barrios, & Holguin, 2002), but they are far less modifiable with patients. With a view to the possible introduction of rescripting into CBT for health anxiety, preliminary research has shown that rescripting can significantly reduce the distressing reactions to the images (Philips & Samson, in preparation). The mere repetition of the formation of the index images did not reduce the emotional reactions. The long term effects of rescripting exercises remain to be determined. These recent findings on intrusive images in health anxiety, and in pain, are promising, and mightmake a significant contribution to the enhancement of cognitive behaviour therapy for health anxiety (Hackmann et al., 2011). Cognitive biases The operation of various cognitive biases can sustain and reinforce health anxiety. For example, ex consequentia reasoning ( Arntz et al., 1995), a bias that arises from the usually latent belief that, ‘If I am anxious it must mean that there is danger; so, if I am made anxious by the pain in my head that means there is a significant danger lurking’, can promote health anxiety Similarly, confirmatory biases can reinforce the anxiety. “As a result” of confirmatory biases “patients selectively notice and remember information consistent with negative beliefs about their problems” ( Salkovskis, 1996, p. 69). One of the two thought–action–fusion (TAF) biases, inflated likelihood of occurrence, can play a part in health anxiety. In this bias the affected person believes that thoughts of a particular misfortune, such as a thought about developing cancer, actually increase the probability that the misfortune will occur: the likelihood form of TAF (Shafran, Thordarson, & Rachman, 1996). Recently it has been suggested that intrusive images are also subject to the likelihood bias. Philips (2011) encountered ‘image–action–fusion’ in a rehabilitation sample of sufferers from chronic pain. Many participants endorsed a belief that having their recurrent, unwanted intrusive images about the pain and its disabling consequences, increased the probability that the dreaded consequence would actually occur. Over-predictions of the intensity of an anticipated fear, a robust phenomenon, is another bias that can increase health anxiety ( Rachman & Bichard, 1988). Safety behaviour Safety behaviour is a prominent feature of HAD, as it is in other anxiety disorders, and is almost universally regarded as undesirable. Sufferers from HAD attempt to cope with threats to their health by carrying out safety behaviours that are emblematic of the disorder. Prominent forms of safety behaviour are avoidance (of hospitals, specialist clinics, sick people, blood, mentally ill people, and so forth), repeated medical consultations and tests, self-checking (of one's body, one's memory, vision), compulsive requests for reassurance, repeated searches for reassuring information, distraction, thought suppression, excessive cleaning, excessive concern with food preparation and dieting. There is ample evidence that safety behaviour can reinforce unadaptive behaviour, and interfere with the progress of therapy (Rachman et al., 2008). In a well conceived experiment Olatunji, Etzel, Tomarken, Cieselski, and Deacon (2011) demonstrated that engaging in hygienic safety behaviours, such as repeatedly using sanitized wipes, exacerbated ‘symptoms’ of health anxiety, and increased scores on measures of HAD. Safety behaviour can interfere with a prevailing treatment technique, ERP, because it disrupts or undermines the effects of the exposure exercises, and in CBT a resort to safety behaviour can interfere with the required disconfirmations of the maladaptive cognitions. In the longer term safety behaviour is partly responsible for maintaining the anxiety disorder. Recently, the role and effects of safety behaviour were reconsidered because of accumulating evidence that in specifiable circumstances ‘judicious safety behaviour’ can facilitate therapeutic progress (Rachman et al., 2008). It has been shown to be facilitative in ERP treatment of agoraphobia (Rachman, Craske, Tallman, & Solyom, 1986; de Silva & Rachman, 1984), in the treatment of snake phobia (Bandura, Jeffery, & Wright, 1974; Milosevic & Radomsky, 2008; Rachman, Hammond, & Radomsky, 2000), in treating acrophobia (Ritter, 1969), fear of spiders (Hood, Antony, Koerner, & Monson, 2010), reducing claustrophobia (Powers, Smits, & Telch, 2004; Sy, Dixon, Lickel, Nelson, & Deacon, 2011) and in reducing feelings of contamination, disgust and anxiety in non-clinical participants who react strongly to contact with a contaminant (Rachman, Shafran, Radomsky, & Zysk, 2011; Van den Hout, Engelhard, Toffolo & van Uijen, 2011). In none of these experiments was there evidence that the use of the safety behaviour prevented the desired changes in maladaptive cognitions. Increasing attention is being paid to re-considering the effects of safety behaviour (Parrish & Radomsky, 2010; Parrish, Radomsky, & Dugas, 2008; Sy et al., 2011). Just as there is bad cholesterol and good cholesterol, there is bad safety behaviour and good safety behaviour. In HAD there are four main forms of safety behaviour: avoidance, checking, information seeking and requests for reassurance. None of them appear to be helpful. The fear of serious illness raises the person's level of vigilance which is manifested in scanning for internal symptoms and signs discernible on one's body. The external environment is scanned for items, situations, people that are potential threats to one's health, and when detected they increase the person's anxiety and hence initiate avoidance behaviour. The avoidance behaviour precludes acquiring reassuring information about the perceived threat, and does nothing to disconfirm the maladaptive cognitions. The fundamental overestimations of the probability of becoming ill and of the probability of the illness being very serious, persist. Many affected people engage in repeated and increasingly intense checking of their bodies and bodily functions. Regrettably the checking generally increases their anxiety partly because of the enhanced attention paid to the seemingly suspicious signs and symptoms, and because repeated checking tends to produce a loss of confidence in one's memory (Radomsky, Rachman, & Hammond, 2001). Compulsive checking cause more checking, not less (Radomsky, Shafran, Coughtrey, & Rachman, 2010). A clear association between health anxiety and safety behaviour was found in an experiment on 20 chronic pain patients with high health anxiety and 20 with low health anxiety. Tang et al. (2007) found that the participants with high anxiety engaged in more safety behaviours than did those with low anxiety, and especially interesting for the cognitive model, there was a correlation between catastrophizing thoughts about the pain and safety behaviour. Repeated requests for reassurance are a characteristic feature of health anxiety, and can be frustrating and exasperating. They are generally unsuccessful and can be an impediment to treatment. It is an ineffective form of safety behaviour (Parrish & Radomsky, 2010). The development of an acceptable and effective form of reassurance, perhaps following along the lines of ‘judicious safety behaviour’, would be a significant forward step. Requests for reassurance, direct or indirect, tend to provide some satisfaction and relief but the effects are so transient that the requests are repeated many times. The reassurances appear to have little or no effect on the person's overestimations of the probability and of the seriousness of the feared illness. Meechan, Collins, Moss-Morrison, and Petrie (2005) found that 33% of the women who were informed that the imaging tests for breast cancer showed no malignant growth were not reassured; as a group they had elevated levels of health anxiety. Another question arises from the fact that only a limited number of selected people can provide sufferers with the responses that provide relief, albeit transient. Some of the qualities of the satisfactory sources of reassurance are obvious (sympathetic, patient, authoritative) but others are idiosyncratic; their responses have the elusive ‘just right’ properties that many patients with obsessive compulsive disorders strive to achieve. These questions are best addressed by a re-construal of the nature and purpose of the compulsive requests for reassurance. The sufferer is apparently seeking information (e.g., ‘Is this blemish on my face cancerous?’) but knows in advance what the answer will be. When the question is repeated an hour later, they still know what the answer will be. The sufferer is not impeded by a deficit of memory. The person is seeking some relief from distress and anxiety, not information as such. Without necessarily recognising it themselves, sufferers are seeking a response not information as such. Spurts of anxiety recur and these generate a search for relief, for reassurance; so the question is asked again and again, usually put to a reliably sympathetic and patient listener. This re-construal of repetitive reassurance-seeking as an attempt to gain relief from anxiety, opens the way to shaping judicious safety behaviours that can provide an enhanced and lasting sense of safety about the perceived threats to one's health. The use of judicious safety behaviour might prove to be useful addition to the treatment of HAD (Rachman et al., 2008). In CBT the anxiety-elevating effects of random excessive searching is reduced by changes in cognition. People are encouraged to become selective and obtain information from authoritative people and sources. Specific recommendations are made after determining the person's current behaviour and what they expect to gain from their searches (What information is helpful to you? Can it help you better to protect yourself? Does it make you less anxious? Has an unreliable source ever made you more concerned and anxious about your health? Who and what are reliable sources, and how can you avoid the unreliable ones?). HAD and panic disorder There are important connections between health anxiety and panic disorder. The first is a similarity in certain features of both disorders and the second is that the prevailing explanations of health anxiety and panic disorder share an essential cognitive element. In both theories it is postulated that the anxiety and fear are caused by a catastrophic misinterpretation of certain bodily sensations (Clark, 1986; Salkovskis & Warwick, 2001). The experience of intensified autonomic sensations, such as a pounding heart, can evoke anxiety. If these sensations are misinterpreted as signs of a catastrophic threat, a panic can occur. However, in health anxiety the potentially threatening sensations are wider, and the triggers for anxiety include images, thoughts, loss of energy, perceived loss of memory, recurrent intrusive thoughts, and other changes in functioning or appearance. In panic disorder the misinterpretation of the bodily sensations provokes an intense fear of imminent danger, often of a heart attack (Rachman & de Silva, 2010a). The fear involves autonomic arousal, arises sharply, lasts in the region of 5–20 min, and gradually subsides only to leave a residue of anxiety. The episodes of panic tend to occur in particular situations such as public places, and the disorder is associated with agoraphobia. In contrast, the misinterpretations of bodily sensations or functions, disturbing thoughts, intrusive images and so on, that provoke health anxiety create a persisting well of inflated anxiety that generates pervasive hypervigilance. The cognitive content of health anxiety is circumscribed (e.g., AIDS) or general, a dread of a range of possible illnesses. Some patients are tormented by both types of cognition.